2 Collate Data

Purpose: Collate EcoDrought streamflow and temperature data, povided by EcoD PIs/partners and NWIS

Notes:

need to figure out how to embed shiny/interactive plots. Tried ‘shinylive’ but this does not work, and appears to be a known issue (instability) with the latest Quarto release (v1.6.32)
- here is the shinylive demo https://github.com/coatless-quarto/r-shinylive-demo
- see https://github.com/coatless-quarto/r-shinylive-demo/issues/14 which suggests downgrading
need to address missing data issue in the Flathead…interpolate at hourly timescale?

2.1 Gather site information

Code

# West Brook
siteinfo_wb <- read_csv("C:/Users/jbaldock/OneDrive - DOI/Documents/USGS/EcoDrought/EcoDrought Working/Data/Raw data/Mass/MA_site_info.csv") %>%
  mutate(Station_No = factor(Station_No), Site_Name = factor(Site_Name)) %>%
  rename(site_id = Site_ID, station_no = Station_No, site_name = Site_Name, lat = Latitude_dec_deg, long = Longitude_dec_deg, elev_ft = Elevation_ft, area_sqmi = Drainage_Area_sqmi) %>%
  mutate(designation = "little", basin = "West Brook", region = "Mass") #%>% select(-c(elev_ft, area_sqmi)) 

# Shenandoah
siteinfo_shen <- read_csv("C:/Users/jbaldock/OneDrive - DOI/Documents/USGS/EcoDrought/EcoDrought Working/Data/Raw data/Virg/VA_site_info.csv") %>%
  mutate(Station_No = factor(Station_No), Site_Name = factor(Site_Name)) %>%
  rename(site_id = Site_ID, station_no = Station_No, site_name = Site_Name, lat = Latitude_dec_deg, long = Longitude_dec_deg, elev_ft = Elevation_ft, area_sqmi = Drainage_Area_sqmi) %>%
  mutate(designation = ifelse(str_detect(site_id, "10FL"), "big", "little"), 
         basin = str_sub(site_name, 1, str_length(site_name)-3), region = "Shen") #%>% select(-c(elev_ft, area_sqmi))

# Flathead/Muhlfeld
siteinfo_flat <- read_csv("C:/Users/jbaldock/OneDrive - DOI/Documents/USGS/EcoDrought/EcoDrought Working/Data/Raw data/Flathead/Flathead_SiteInfo_UpdateOct25.csv") %>% 
  select(basin, site_name, site_id, region, designation, lat, long) %>%
  rename(subbasin = basin) %>%
  mutate(basin = "Flathead", region = "Flat") %>%
  select(site_id, site_name, lat, long, designation, basin, subbasin, region) %>%
  filter(designation == "little")
  
# GYA/Al-Chokhachy
siteinfo_gya <- read_csv("C:/Users/jbaldock/OneDrive - DOI/Documents/USGS/EcoDrought/EcoDrought Working/Data/Raw data/Al-Chokhachy/Al-Chokhachy_sites.csv") %>%
  mutate(region = ifelse(basin == "Snake River", "Snake", "Shields"), 
         designation = "little") %>%
  select(site_id, site_name, latitude, longitude, designation, basin, region) %>% 
  rename(lat = latitude, long = longitude)
  

# NWIS Medium/Big/Super G
sites <- c("01169900", # South River at Conway, Massachusetts
           "13011500", # Pacific Creek, Snake River, Wyoming
           "06195600", # Shields River at Livingston, Montana
           "12355500", # North Fork Flathead River, Montana
           "10396000", # Donner Blitzen River near Frenchglen, Oregon
           # Medium G
           "12355347", # Big Creek (Flathead)
           "12355342", # Hallowat Creek (Flathead)
           "06192980", # Shields Rivera above Smith Creek (GYA)
           "06192900", # Dugout Creek Mouth (GYA)
           "13012475", # South Fork Spread Creek (GYA)
           "13012465", # Leidy Creek, lower (GYA)
           "01171100", # West Brook (Mass)
           "01171000",  # Avery Brook (Mass)
           "424551118503200", # Fish Creek at DB confluence (Oreg)
           "424547118503500", # DB above Fish Creek (Oreg)
           "424325118495900", # DB near Burnt Car Spring (Oreg)
           "424003118453700", # Little Blitzen River (Oreg)
           "423830118453200", # Indian Creek (Oreg)
           "423815118453900" # DB above Indian Creek (Oreg)
           )
siteinfo_nwis <- tibble(readNWISsite(sites)[,c(2:3,7,8,20,30)]) # get site info
names(siteinfo_nwis) <- c("station_no", "site_name", "lat", "long", "elev_ft", "area_sqmi") # rename columns
siteinfo_nwis <- siteinfo_nwis %>% mutate(site_name = c("South River Conway NWIS", 
                                                        "Avery Broook NWIS", 
                                                        "West Brook NWIS", 
                                                        "Dugout Creek NWIS", 
                                                        "Shields River ab Smith NWIS", 
                                                        "Shields River nr Livingston NWIS", 
                                                        "Donner Blitzen River nr Frenchglen NWIS", 
                                                        "Hallowat Creek NWIS", 
                                                        "Big Creek NWIS", 
                                                        "North Fork Flathead River NWIS", 
                                                        "Pacific Creek at Moran NWIS", 
                                                        "Leidy Creek Mouth NWIS", 
                                                        "SF Spread Creek Lower NWIS",
                                                        "Donner Blitzen ab Indian NWIS",
                                                        "Indian Creek NWIS",
                                                        "Little Blizten River NWIS",
                                                        "Donner Blitzen nr Burnt Car NWIS",
                                                        "Donner Blitzen ab Fish NWIS",
                                                        "Fish Creek"),
                                          site_id = c("SRC", "AVB", "WBR", "DUG", "SRS", "SRL", "DBF", "HAL", "BIG", "NFF", "PCM", "LEI", "SFS", "DBI", "IND", "LBL", "DBB", "DBA", "FSH"),
                                          designation = c("big", "medium", "medium", "medium", "medium", "big", "big", "medium", "medium", "big", "big", "medium", "medium", "medium", "medium", "medium", "medium", "medium", "medium"),
                                          basin = c("West Brook", "West Brook", "West Brook", "Shields River", "Shields River", "Shields River", "Donner Blitzen", "Flathead", "Flathead", "Flathead", "Snake River", "Snake River", "Snake River", "Donner Blitzen", "Donner Blitzen", "Donner Blitzen", "Donner Blitzen", "Donner Blitzen", "Donner Blitzen"),
                                          region = c("Mass", "Mass", "Mass", "Shields", "Shields", "Shields", "Oreg", "Flat", "Flat", "Flat", "Snake", "Snake", "Snake","Oreg", "Oreg", "Oreg", "Oreg", "Oreg", "Oreg")) %>% 
  select(site_id, site_name, lat, long, station_no, designation, basin, region, elev_ft, area_sqmi)
#mapview(st_as_sf(siteinfo_nwis, coords = c("long", "lat"), crs = 4326))


# bind together, fill in ragged subbasin
siteinfo <- bind_rows(siteinfo_wb, siteinfo_shen, siteinfo_flat, siteinfo_gya, siteinfo_nwis)
siteinfo$subbasin[siteinfo$site_name == "Hallowat Creek NWIS"] <- "Big Creek"
siteinfo$subbasin[siteinfo$site_name == "Big Creek NWIS"] <- "Big Creek"
siteinfo <- siteinfo %>% mutate(subbasin = ifelse(is.na(subbasin), basin, subbasin))


# fix Shields River Valley Ranch site locations
siteinfo$lat[siteinfo$site_id == "SH07"] <- siteinfo$lat[siteinfo$site_id == "SRS"]
siteinfo$long[siteinfo$site_id == "SH07"] <- siteinfo$long[siteinfo$site_id == "SRS"]


# add elevation and area variables (from watershed delineation)
areafiles <- list.files("C:/Users/jbaldock/OneDrive - DOI/Documents/USGS/EcoDrought/EcoDrought Working/EcoDrought-Analysis/Watershed Delineation/Area and Elevation")
arealist <- list()
for (i in 1:length(areafiles)) { arealist[[i]] <- read_csv(paste("C:/Users/jbaldock/OneDrive - DOI/Documents/USGS/EcoDrought/EcoDrought Working/EcoDrought-Analysis/Watershed Delineation/Area and Elevation/", areafiles[i], sep = ""))}
areaelev <- do.call(rbind, arealist)
# how well do provided and delineation area/elevation match?
# siteinfo %>% left_join(areaelev, by = "site_id") %>% ggplot() + geom_point(aes(x = area_sqmi.x, y = area_sqmi.y)) + geom_abline(intercept = 0, slope = 1) + facet_wrap(~basin, scales = "free")
# test %>% ggplot() + geom_point(aes(x = elev_ft.x, y = elev_ft.y)) + geom_abline(intercept = 0, slope = 1) + facet_wrap(~basin, scales = "free")
# add delineated variables
siteinfo <- siteinfo %>% select(-c(area_sqmi, elev_ft)) %>% left_join(areaelev)
# fix NF Flathead (no dem from Canada)
siteinfo$area_sqmi[siteinfo$site_id == "NFF"] <- 1556

Write and re-load site information

Code

write_csv(siteinfo, "C:/Users/jbaldock/OneDrive - DOI/Documents/USGS/EcoDrought/EcoDrought Working/Data/EcoDrought_SiteInformation.csv")
siteinfo <- read_csv("C:/Users/jbaldock/OneDrive - DOI/Documents/USGS/EcoDrought/EcoDrought Working/Data/EcoDrought_SiteInformation.csv")

View unique basin names

Code

unique(siteinfo$basin)

[1] "West Brook"     "Paine Run"      "Piney River"    "Staunton River"
[5] "Flathead"       "Duck Creek"     "Shields River"  "Snake River"   
[9] "Donner Blitzen"

View unique site names

Code

unique(siteinfo$site_name)

  [1] "Avery Brook"                            
  [2] "Jimmy Brook"                            
  [3] "Mitchell Brook"                         
  [4] "Obear Brook Lower"                      
  [5] "Sanderson Brook"                        
  [6] "West Brook Lower"                       
  [7] "West Brook Upper"                       
  [8] "West Brook Reservoir"                   
  [9] "West Whately Brook"                     
 [10] "West Brook 0"                           
 [11] "Paine Run 01"                           
 [12] "Paine Run 02"                           
 [13] "Paine Run 03"                           
 [14] "Paine Run 04"                           
 [15] "Paine Run 05"                           
 [16] "Paine Run 06"                           
 [17] "Paine Run 07"                           
 [18] "Paine Run 08"                           
 [19] "Paine Run 09"                           
 [20] "Paine Run 10"                           
 [21] "Piney River 01"                         
 [22] "Piney River 02"                         
 [23] "Piney River 03"                         
 [24] "Piney River 04"                         
 [25] "Piney River 05"                         
 [26] "Piney River 06"                         
 [27] "Piney River 08"                         
 [28] "Piney River 09"                         
 [29] "Piney River 10"                         
 [30] "Staunton River 01"                      
 [31] "Staunton River 02"                      
 [32] "Staunton River 03"                      
 [33] "Staunton River 04"                      
 [34] "Staunton River 05"                      
 [35] "Staunton River 06"                      
 [36] "Staunton River 07"                      
 [37] "Staunton River 08"                      
 [38] "Staunton River 09"                      
 [39] "Staunton River 10"                      
 [40] "BigCreekLower"                          
 [41] "BigCreekMiddle"                         
 [42] "BigCreekUpper"                          
 [43] "CoalCreekHeadwaters"                    
 [44] "CoalCreekLower"                         
 [45] "CoalCreekMiddle"                        
 [46] "CycloneCreekLower"                      
 [47] "CycloneCreekMiddle"                     
 [48] "CycloneCreekUpper"                      
 [49] "HallowattCreekLower"                    
 [50] "LangfordCreekLower"                     
 [51] "LangfordCreekUpper"                     
 [52] "McGeeCreekLower"                        
 [53] "McGeeCreekTrib"                         
 [54] "McGeeCreekUpper"                        
 [55] "MeadowCreek"                            
 [56] "NicolaCreek"                            
 [57] "CoalCreekNorth"                         
 [58] "SkookoleelCreek"                        
 [59] "WernerCreek"                            
 [60] "WoundedBuckCreek"                       
 [61] "EF Duck Creek be HF"                    
 [62] "EF Duck Creek ab HF"                    
 [63] "Henrys Fork"                            
 [64] "Brackett Creek"                         
 [65] "Buck Creek"                             
 [66] "Crandall Creek"                         
 [67] "Deep Creek"                             
 [68] "Dugout Creek"                           
 [69] "Lodgepole Creek"                        
 [70] "Shields River Valley Ranch"             
 [71] "Shields River ab Dugout"                
 [72] "Grizzly Creek"                          
 [73] "Grouse Creek"                           
 [74] "Leidy Creek Lower"                      
 [75] "Leidy Creek Upper"                      
 [76] "Leidy Creek Mouth"                      
 [77] "NF Spread Creek Lower"                  
 [78] "NF Spread Creek Upper"                  
 [79] "Rock Creek"                             
 [80] "SF Spread Creek Lower"                  
 [81] "SF Spread Creek Upper"                  
 [82] "Spread Creek Dam"                       
 [83] "South River Conway NWIS"                
 [84] "Avery Broook NWIS"                      
 [85] "West Brook NWIS"                        
 [86] "Dugout Creek NWIS"                      
 [87] "Shields River ab Smith NWIS"            
 [88] "Shields River nr Livingston NWIS"       
 [89] "Donner Blitzen River nr Frenchglen NWIS"
 [90] "Hallowat Creek NWIS"                    
 [91] "Big Creek NWIS"                         
 [92] "North Fork Flathead River NWIS"         
 [93] "Pacific Creek at Moran NWIS"            
 [94] "Leidy Creek Mouth NWIS"                 
 [95] "SF Spread Creek Lower NWIS"             
 [96] "Donner Blitzen ab Indian NWIS"          
 [97] "Indian Creek NWIS"                      
 [98] "Little Blizten River NWIS"              
 [99] "Donner Blitzen nr Burnt Car NWIS"       
[100] "Donner Blitzen ab Fish NWIS"            
[101] "Fish Creek"

Map sites

Code

# convert to spatial object and view on map
#| fig-cap: "Map of EcoDrought project locations"
siteinfo_sp <- st_as_sf(siteinfo, coords = c("long", "lat"), crs = 4326)
mapview(siteinfo_sp, zcol = "designation")

2.2 Load EcoD data

Code

# West Brook
dat_wb <- read_csv("C:/Users/jbaldock/OneDrive - DOI/Documents/USGS/EcoDrought/EcoDrought Working/Data/Raw data/Mass/EcoDrought Continuous_MA.csv") %>%
  mutate(Station_No = factor(Station_No), Site_Name = factor(Site_Name)) %>%
  rename(station_no = Station_No, site_name = Site_Name, datetime = DateTime_EST, height = GageHeight_Hobo_ft, flow = Discharge_Hobo_cfs, tempf = WaterTemperature_HOBO_DegF) %>%
  mutate(tempc = (tempf - 32)*(5/9)) %>% select(station_no, site_name, datetime, height, flow, tempc) %>%
  left_join(siteinfo %>% select(-station_no))


# Shenandoah
dat_shen <- read_csv("C:/Users/jbaldock/OneDrive - DOI/Documents/USGS/EcoDrought/EcoDrought Working/Data/Raw data/Virg/EcoDrought_Continuous_VA.csv") %>%
  mutate(Station_No = factor(Station_No), Site_ID = factor(Site_ID), Discharge_Hobo_cfs = as.numeric(Discharge_Hobo_cfs)) %>%
  rename(station_no = Station_No, site_id = Site_ID, datetime = DateTime_EST, height = GageHeight_Hobo_ft, flow = Discharge_Hobo_cfs, tempf = WaterTemperature_HOBO_DegF) %>%
  mutate(tempc = (tempf - 32)*(5/9)) %>% select(station_no, site_id, datetime, height, flow, tempc) %>%
  left_join(siteinfo)
# pull in Big G data separately (UVA long-term gage sites)
dat_shen_uva_q <- read_csv("C:/Users/jbaldock/OneDrive - DOI/Documents/USGS/EcoDrought/EcoDrought Working/Data/Raw data/Virg/Shen_BigG_Discharge_hourly_UVA.csv") %>%
  rename(flow = cfs)
dat_shen_uva_t <- read_csv("C:/Users/jbaldock/OneDrive - DOI/Documents/USGS/EcoDrought/EcoDrought Working/Data/Raw data/Virg/Shen_BigG_TempEtc_hourly_UVA.csv") %>%
  select(site_id, datetime, tempc_mean) %>% rename(tempc = tempc_mean)
dat_shen_uva <- dat_shen_uva_q %>% left_join(dat_shen_uva_t) %>% left_join(siteinfo)
# bind usgs and uva data
dat_shen <- bind_rows(dat_shen, dat_shen_uva)


# Flathead/Muhlfeld
flatfiles <- list.files("C:/Users/jbaldock/OneDrive - DOI/Documents/USGS/EcoDrought/EcoDrought Working/Data/Raw data/Flathead/Export2/EMA/Continuous Data")
flatlist <- list()
for (i in 1:length(flatfiles)) { 
  #print(flatfiles[i])
  flatlist[[i]] <- read_csv(paste("C:/Users/jbaldock/OneDrive - DOI/Documents/USGS/EcoDrought/EcoDrought Working/Data/Raw data/Flathead/Export2/EMA/Continuous Data/", flatfiles[i], sep = "")) %>%
    mutate(DateTime = mdy_hm(DateTime, tz = "MST"),
           site_name = gsub("EMA.csv", "", flatfiles[i]))
  }
dat_flat <- bind_rows(flatlist) %>% select(DateTime, GageHeightFT, DischargeCFS, TempF, TempC, site_name, DischargeReliability, TempReliability) %>% 
  rename(datetime = DateTime, height = GageHeightFT, flow = DischargeCFS, tempf = TempF, tempc = TempC) %>%
  mutate(DischargeReliability = as_factor(DischargeReliability),
         TempReliability = as_factor(TempReliability)) %>%
  mutate(tempf = ifelse(is.na(tempf), (tempc * (9/5)) + 32, tempf),
         tempc = ifelse(is.na(tempc), (tempf - 32) * (5/9), tempc)) %>%
  left_join(siteinfo) %>% select(-tempf)


# Greater Yellowstone/Al-Chokhachy
gyafiles <- list.files("C:/Users/jbaldock/OneDrive - DOI/Documents/USGS/EcoDrought/EcoDrought Working/Data/Raw data/Al-Chokhachy/Al-Chokhachy data files")
gyalist <- list()
for (i in 1:length(gyafiles)) { 
  #print(gyafiles[i])
  gyalist[[i]] <- read_csv(paste("C:/Users/jbaldock/OneDrive - DOI/Documents/USGS/EcoDrought/EcoDrought Working/Data/Raw data/Al-Chokhachy/Al-Chokhachy data files/", gyafiles[i], sep = "")) %>%
    mutate(date = mdy(date), DateTime = ymd_hms(paste(date, time, sep = " "), tz = "MST"), discharge = as.numeric(discharge)*35.314666212661)
}
dat_gya <- bind_rows(gyalist) %>% select(DateTime, depth, discharge, temperature, location) %>% 
  rename(datetime = DateTime, height = depth, flow = discharge, tempc = temperature, site_name = location) %>%
  filter(site_name != "EF Henrys") %>% # drop weird duplicate site/year?
  mutate(site_name = dplyr::recode(site_name,
                            "EF Above Confluence" = "EF Duck Creek ab HF",
                            "EF Below Confluence" = "EF Duck Creek be HF",
                            "NF Spread Creek" = "NF Spread Creek Lower",
                            "Upper NF Spread Creek" = "NF Spread Creek Upper",
                            "SF Spread Creek" = "SF Spread Creek Lower",
                            "Upper SF Spread Creek" = "SF Spread Creek Upper",
                            "Shields River above Dugout Creek" = "Shields River ab Dugout",
                            "Upper Leidy Creek" = "Leidy Creek Upper", 
                            "Leidy Creek" = "Leidy Creek Mouth",
                            "Spread Creek" = "Spread Creek Dam",
                            "Shields River above Smith Creek" = "Shields River Valley Ranch")) %>%
  left_join(siteinfo) %>% filter(tempc <= 100)

Bind EcoD hourly flow/temp data with siteinfo and write to file

Code

# bind together
dat <- bind_rows(dat_wb, dat_shen, dat_flat, dat_gya)
# unique(dat$site_name)
write_csv(dat, "C:/Users/jbaldock/OneDrive - DOI/Documents/USGS/EcoDrought/EcoDrought Working/Data/EcoDrought_FlowTempData_Raw.csv")
dat <- read_csv("C:/Users/jbaldock/OneDrive - DOI/Documents/USGS/EcoDrought/EcoDrought Working/Data/EcoDrought_FlowTempData_Raw.csv")

Check unique designations

Code

unique(dat$designation)

[1] "little" "big"

Check unique regions

Code

unique(dat$region)

[1] "Mass"    "Shen"    "Flat"    "Shields" "Snake"

Check unique basins

Code

unique(dat$basin)

[1] "West Brook"     "Paine Run"      "Staunton River" "Piney River"   
[5] "Flathead"       "Shields River"  "Duck Creek"     "Snake River"

Check unique subbasins

Code

unique(dat$subbasin)

 [1] "West Brook"         "Paine Run"          "Staunton River"    
 [4] "Piney River"        "Big Creek"          "Coal Creek"        
 [7] "McGee Creek"        "Wounded Buck Creek" "Shields River"     
[10] "Duck Creek"         "Snake River"

2.2.1 Inspect hourly data

Code

mysites <- unlist(unique(dat %>% filter(basin == "Flathead") %>% select(site_name)))
myplots <- list()
for (i in 1:length(mysites)) {
  test <- dat %>% filter(site_name == mysites[i]) %>% select(datetime, flow, DischargeReliability) %>% spread(key = DischargeReliability, value = flow) %>% arrange(datetime) %>% mutate(mindiff = as.numeric(lead(datetime) - (datetime)))
na_index <- which(test$mindiff > 60)
  ds_blank <- tibble(index = na_index + 0.5, datetime = NA, "0" = NA, "1" = NA)
  myplots[[i]] <- test %>% select(-mindiff) %>% rowid_to_column("index") %>% union(ds_blank) %>% arrange(index) %>% select(-index) %>% mutate(datetime = if_else(is.na(datetime), lag(datetime) + hours(1), datetime)) %>% dygraph() %>% dyRangeSelector() %>% dyAxis("y", label = "flow (cfs)") %>% dyOptions(fillGraph = TRUE, drawGrid = FALSE, axisLineWidth = 1.5)
}

Code

dat %>% filter(site_name == "CoalCreekLower") %>% select(datetime, flow, DischargeReliability) %>% spread(key = DischargeReliability, value = flow) %>% dygraph() %>% dyRangeSelector() %>% dyAxis("y", label = "flow (cfs)") %>% dyOptions(fillGraph = TRUE, drawGrid = FALSE, axisLineWidth = 1.5)

2.3 Load NWIS data

Code

# extract daily mean discharge and temp data from USGS NWIS website
dat_superg_nwis <- tibble(readNWISdv(siteNumbers = sites, parameterCd = c("00010", "00060"), startDate = "1980-01-01", endDate = Sys.Date(), statCd = c("00003", "00001", "00002")))[,-c(1,5,7,9,11)]
names(dat_superg_nwis) <- c("station_no", "date", "tempc_max", "tempc_min", "tempc_mean", "flow_mean")
dat_superg_nwis <- dat_superg_nwis %>% left_join(siteinfo)

# Manually grab Donner Blitzen above Indian Creek at the original timescale
# daily mean flow dropped from above query, perhaps b/c in 2020 flow measurements were made every minute
dbabind <- tibble(readNWISdata(sites = "423815118453900", service = "uv", startDate = "1980-01-01", endDate = Sys.Date()))[,c(2,3,6)] 
dbabind_daily <- dbabind %>% group_by(site_no, date(dateTime)) %>% summarize(nobs = n(), flow_mean = mean(X_00060_00000)) %>% filter(nobs %in% c(96, 1440)) %>% select(c(1,2,4))
names(dbabind_daily) <- c("station_no", "date", "flow_mean") 
dbabind_daily <- dat_superg_nwis %>% filter(station_no == "423815118453900") %>% select(-flow_mean) %>% left_join(dbabind_daily)

# bind DBabInd to reset of data
dat_superg_nwis <- rbind(dat_superg_nwis %>% filter(station_no != "423815118453900"), dbabind_daily)

Code

dat_superg_nwis %>% filter(designation == "big") %>% ggplot() + geom_line(aes(x = date, y = log(flow_mean))) + facet_wrap(~site_name)

(log) Stream flow (cfs) for Big G NWIS gages

Code

dat_superg_nwis %>% filter(designation == "medium") %>% ggplot() + geom_line(aes(x = date, y = log(flow_mean))) + facet_wrap(~site_name)

(log) Stream flow (cfs) for Medium G NWIS gages

2.4 Bind data and calc. daily means

Code

# daily flow/temp summaries
dat_daily <- dat %>% mutate(date = as_date(datetime)) %>% 
  group_by(station_no, site_name, site_id, basin, subbasin, region, lat, long, elev_ft, area_sqmi, designation, date) %>% 
  summarize(disch_reli = max(DischargeReliability),
            temp_reli = max(TempReliability),
            flow_mean = mean(flow), flow_min = min(flow), flow_max = max(flow),
            tempc_mean = mean(tempc), tempc_min = min(tempc), tempc_max = max(tempc)) %>%
  arrange(region, basin, site_name, date) %>%
  ungroup()

# cbind EcoD and NWIS datasets
dat_daily <- bind_rows(dat_daily %>% select(-flow_min, -flow_max, -tempc_min, -tempc_max), 
                       dat_superg_nwis %>% select(-tempc_max, -tempc_min),
                       dat_superg_nwis %>% filter(site_id == "SRL") %>% mutate(subbasin = "Duck Creek") %>% select(-tempc_max, -tempc_min))

# add missing dates
dat_daily <- fill_missing_dates(dat_daily, dates = date, groups = site_name)

View daily mean time series data, for example, site_name = CoalCreekLower. Notice the many shorts gaps in the daily data.

Code

library(dygraphs)
dat_daily %>% filter(site_name == "CoalCreekLower") %>% select(date, flow_mean) %>% dygraph() %>% dyRangeSelector() %>% dyAxis("y", label = "Mean daily flow (cfs)") %>% dyOptions(fillGraph = TRUE, drawGrid = FALSE, axisLineWidth = 1.5)

2.5 Interpolate missing data

Small periods of missing data (<24 hours) become a problem when aggregating at the daily and weekly time scales. *Note that currently this discovers and fills missing data at the daily time scale, but should be changed to interpolate at the original timescale of the raw data (e.g., hourly).

Code

# explore data gaps
mysites <- unique(dat_daily$site_name)
mynas <- list()
for (i in 1:length(mysites)) {
  mydisch <- unlist(dat_daily$flow_mean[dat_daily$site_name == mysites[i]])
  runsna <- rle(is.na(mydisch))
  mynas[[i]] <- tibble(site_name = mysites[i], run = runsna$lengths[runsna$values == TRUE])
}
mynas <- do.call(rbind, mynas)
# mynas %>% filter(run < 100) %>% ggplot() + geom_histogram(aes(x = run)) + facet_wrap_paginate(~site_name, nrow = 4, ncol = 4, page = 1)
# mynas %>% filter(run < 100) %>% ggplot() + geom_histogram(aes(x = run)) + facet_wrap_paginate(~site_name, nrow = 4, ncol = 4, page = 2)
# mynas %>% filter(run < 100) %>% ggplot() + geom_histogram(aes(x = run)) + facet_wrap_paginate(~site_name, nrow = 4, ncol = 4, page = 3)
# mynas %>% filter(run < 100) %>% ggplot() + geom_histogram(aes(x = run)) + facet_wrap_paginate(~site_name, nrow = 4, ncol = 4, page = 4)

Most gaps are relatively short

Code

mynas %>% ggplot() + geom_histogram(aes(x = run)) + xlab("Days") + ylab("Frequency")

Distributiion of lengths of missing data (days)

Zoomed in…

Code

mynas %>% filter(run < 30) %>% ggplot() + geom_histogram(aes(x = run))  + xlab("Days") + ylab("Frequency")

Considering just the short gaps, which are likely a function of logger malfunction or Aquarius export issues, which sites are problematic? Answer: Flathead

Code

mynas %>% filter(run < 30) %>% ggplot() + geom_bar(aes(site_name))+ theme(axis.text.x = element_text(angle = 90, vjust = 0.5, hjust=1))

Frequency of short (<40 days) data gaps by site

Fill short gaps (<=14 days…2x smoothing period) using time series interpolation

Code

# fill short gaps (<=14 days...2x smoothing period) using time series interpolation
datalist <- list()
for (i in 1:length(mysites)) { datalist[[i]] <- dat_daily %>% filter(site_name == mysites[i]) %>% mutate(flow_mean_filled = fillMissing(flow_mean, max.fill = 14, span = 100)) }
# bind and check 1:1
dat_daily_fill <- do.call(rbind, datalist)
# dat_daily_fill %>% ggplot() + geom_point(aes(x = flow_mean, y = flow_mean_filled)) + facet_wrap(~site_name, scales = "free")

Explore interpolated/filled time series relative to original (daily) data Again, CoalCreekLower as an example. Would like to replace with shiny/interactive app

Code

library(dygraphs)
dat_daily_fill %>% filter(site_name == "CoalCreekLower") %>% select(date, flow_mean, flow_mean_filled) %>% dygraph() %>% dyRangeSelector() %>% dySeries("flow_mean", strokeWidth = 5, color = "black") %>% dySeries("flow_mean_filled", strokeWidth = 1, color = "red") %>% dyAxis("y", label = "Mean daily flow (cfs)") %>% dyOptions(fillGraph = TRUE, drawGrid = FALSE, axisLineWidth = 1.5)

2.6 Calculate yield

Code

# convert cfs and basin area to metric
dat_daily_fill <- dat_daily_fill %>% mutate(flow_mean_cms = flow_mean*0.02831683199881, 
                                            flow_mean_filled_cms = flow_mean_filled*0.02831683199881, 
                                            area_sqkm = area_sqmi*2.58999)

# sites
sites <- unique(dat_daily_fill$site_name)

# site-specific basin area in square km
basinarea <- dat_daily_fill %>% filter(!is.na(site_id)) %>% group_by(site_name) %>% summarize(area_sqkm = unique(area_sqkm))

# calculate yield
yield_list <- list()
for (i in 1:length(sites)) {
  d <- dat_daily_fill %>% filter(site_name == sites[i])
  ba <- unlist(basinarea %>% filter(site_name == sites[i]) %>% select(area_sqkm))
  yield_list[[i]] <-  add_daily_yield(data = d, values = flow_mean_cms, basin_area = as.numeric(ba)) %>% left_join(add_daily_yield(data = d, values = flow_mean_filled_cms, basin_area = as.numeric(ba)) %>% rename(Yield_filled_mm = Yield_mm))
}
dat_daily_fill_wyield <- do.call(rbind, yield_list)

2.7 Calculate 7-day means

Code

dat_daily_fill_wyield <- dat_daily_fill_wyield %>%
  group_by(site_name) %>%
  mutate(flow_mean_7 = rollapply(flow_mean, FUN = mean, width = 7, align = "center", fill = NA),
         flow_mean_filled_7 = rollapply(flow_mean_filled, FUN = mean, width = 7, align = "center", fill = NA),
         tempc_mean_7 = rollapply(tempc_mean, FUN = mean, width = 7, align = "center", fill = NA),
         Yield_mm_7 = rollapply(Yield_mm, FUN = mean, width = 7, align = "center", fill = NA),
         Yield_filled_mm_7 = rollapply(Yield_filled_mm, FUN = mean, width = 7, align = "center", fill = NA)) %>%
  ungroup() %>% filter(!is.na(site_id))

# # view flow
# dat_daily_fill_wyield %>% filter(site_name == "Avery Brook") %>% select(date, flow_mean_filled, flow_mean_filled_7) %>% dygraph() %>% dyRangeSelector() %>% 
#   dySeries("flow_mean_filled", strokeWidth = 5, color = "black") %>% dySeries("flow_mean_filled_7", strokeWidth = 1, color = "red")
# 
# # view yield
# dat_daily_fill_wyield %>% filter(site_name == "Avery Brook") %>% select(date, Yield_filled_mm, Yield_filled_mm_7) %>% dygraph() %>% dyRangeSelector() %>% 
#   dySeries("Yield_filled_mm", strokeWidth = 5, color = "black") %>% dySeries("Yield_filled_mm_7", strokeWidth = 1, color = "red")

# unique(dat_daily_fill_wyield$basin)
# unique(dat_daily_fill_wyield$subbasin)
# unique(dat_daily_fill_wyield$region)
# unique(dat_daily_fill_wyield$disch_reli)
# unique(dat_daily_fill_wyield$temp_reli)

View little and medium g time series data (7-day mean yield), by sub-basin. Use the handles below the x-axis to adjust the time frame.

Code

mysubbasins <- unique(dat_daily_fill_wyield$subbasin)[-c(4,7,13)]
myplots <- list()
for (i in 1:length(mysubbasins)) {
  myplots[[i]] <- dat_daily_fill_wyield %>% filter(designation %in% c("little", "medium"), subbasin == mysubbasins[i]) %>% mutate(logYield = log(Yield_filled_mm_7)) %>% select(date, site_name, logYield) %>% spread(key = site_name, value = logYield) %>% dygraph() %>% dyRangeSelector() %>% dyAxis("y", label = "ln(Yield, mm)")
}

2.8 Write out and read data

Code

# write out
write_csv(dat_daily_fill_wyield, "C:/Users/jbaldock/OneDrive - DOI/Documents/USGS/EcoDrought/EcoDrought Working/Data/EcoDrought_FlowTempData_DailyWeekly.csv")
dat_daily <- read_csv("C:/Users/jbaldock/OneDrive - DOI/Documents/USGS/EcoDrought/EcoDrought Working/Data/EcoDrought_FlowTempData_DailyWeekly.csv")

2.9 View data availability

Code

# summarize data availability
dat_summ <- dat_daily %>% filter(site_id != "MRN") %>%
  group_by(basin, date, designation) %>% summarize(numflow = sum(!is.na(flow_mean)), numtemp = sum(!is.na(tempc_mean))) %>% 
  gather(type, avail, numflow:numtemp) %>% mutate(type2 = as.factor(paste(designation, type, sep = "_"))) %>% 
  mutate(type3 = as.numeric(type2), avail = na_if(avail, 0)) %>% ungroup() %>% filter(!is.na(avail))
# levels(dat_summ$type2)
# unique(dat_summ$basin)

Code

# plot all years
#| fig-cap: "Data availability by basin, all years"
myplot <- dat_summ %>% ggplot(aes(x = date, y = type3)) + 
  geom_errorbarh(aes(xmax = date, xmin = date, color = avail), size = 0.001) +
  scale_color_continuous(trans = "reverse", na.value = "grey60") +
  scale_y_discrete(limits = c("Big G flow", "Big G temp", "Medium g flow", "Medium g temp", "Little g flow", "Little g temp")) + 
  labs(colour = "Number \nof sites") + ylab("") + xlab("Date") + 
  facet_wrap(~ basin, nrow = 4,
             labeller = as_labeller(c(`West Brook` = "West Brook: G = South River Conway (NWIS)", 
                                      `Staunton River` = "Staunton River: G = Staunton River 10 (UVA)",
                                      `Paine Run` = "Paine Run: G = Paine Run 10 (UVA)",
                                      `Piney River` = "Piney River: G = Piney River 10 (UVA)",
                                      `Snake River` = "Snake River: G = Pacific Creek Moran (NWIS)",
                                      `Shields River` = "Shields River: G = Shields River Livingston (NWIS)",
                                      `Flathead` = "NF Flathead River: G = NF Flathead (NWIS)",
                                      `Donner Blitzen` = "Donner Blitzen River: G = DB Frenchglen (NWIS)",
                                      `Duck Creek` = "Duck Creek: G = Shields River Livingston (NWIS)")))
myplot

Code

# jpeg("C:/Users/jbaldock/OneDrive - DOI/Documents/USGS/EcoDrought/EcoDrought Working/EcoDrought-Analysis/Data Availability/DataAvailability_byBasin.jpg", height = 6, width = 12, units = "in", res = 500)
# myplot
# dev.off()

Code

# plot recent years
#| fig-cap: "Data availability by basin, recent years"
# jpeg("./Data Availability/DataAvailability_byBasin_recent.jpg", height = 6, width = 12, units = "in", res = 500)
myplot <- dat_summ %>% filter(date >= "2018-10-01") %>% ggplot(aes(x = date, y = type3)) + 
  geom_errorbarh(aes(xmax = date, xmin = date, color = avail), size = 0.001) +
  scale_color_continuous(trans = "reverse", na.value = "grey60") +
  scale_y_discrete(limits = c("Big G flow", "Big G temp", "Medium g flow", "Medium g temp", "Little g flow", "Little g temp")) + 
  labs(colour = "Number \nof sites") + ylab("") + xlab("Date") + 
  facet_wrap(~ basin, nrow = 4,
             labeller = as_labeller(c(`West Brook` = "West Brook: G = South River Conway (NWIS)", 
                                      `Staunton River` = "Staunton River: G = Staunton River 10 (UVA)",
                                      `Paine Run` = "Paine Run: G = Paine Run 10 (UVA)",
                                      `Piney River` = "Piney River: G = Piney River 10 (UVA)",
                                      `Snake River` = "Snake River: G = Pacific Creek Moran (NWIS)",
                                      `Shields River` = "Shields River: G = Shields River Livingston (NWIS)",
                                      `Flathead` = "NF Flathead River: G = NF Flathead (NWIS)",
                                      `Donner Blitzen` = "Donner Blitzen River: G = DB Frenchglen (NWIS)",
                                      `Duck Creek` = "Duck Creek: G = Shields River Livingston (NWIS)")))
myplot

2.10 Compare co-located gages

2.10.1 Compare synchronous gages

Compare streamflow data from co-located EcoDrought and NWIS gages with overlapping periods of record

Code

# WEST BROOK
p1 <- dat_daily %>% filter(site_name == "West Brook 0") %>% select(date, flow_mean_7) %>% rename(little = flow_mean_7) %>% 
  left_join(dat_daily %>% filter(site_name == "West Brook NWIS") %>% select(date, flow_mean_7) %>% rename(medium = flow_mean_7)) %>%
  ggplot() + geom_point(aes(x = log(medium), y = log(little))) + geom_abline(intercept = 0, slope = 1, color = "red")

p2 <- dat_daily %>% filter(site_name == "Avery Brook") %>% select(date, flow_mean_7) %>% rename(little = flow_mean_7) %>% 
  left_join(dat_daily %>% filter(site_name == "Avery Broook NWIS") %>% select(date, flow_mean_7) %>% rename(medium = flow_mean_7)) %>%
  ggplot() + geom_point(aes(x = log(medium), y = log(little))) + geom_abline(intercept = 0, slope = 1, color = "red")

# FLATHEAD
p3 <- dat_daily %>% filter(site_name == "BigCreekMiddle") %>% select(date, flow_mean_7) %>% rename(little = flow_mean_7) %>% 
  left_join(dat_daily %>% filter(site_name == "Big Creek NWIS") %>% select(date, flow_mean_7) %>% rename(medium = flow_mean_7)) %>%
  ggplot() + geom_point(aes(x = log(medium), y = log(little))) + geom_abline(intercept = 0, slope = 1, color = "red")

# jpeg("./Data Availability/LittleMedium_Co-Located_Gages.jpg", height = 6, width = 6, units = "in", res = 500)
ggarrange(p1, p2, p3, ncol = 2, nrow = 2, labels = c("West Brook 0", "Avery Brook", "Big Creek (Flathead)"))

Streamflow measured at little g gages (EcoDrought) as a function of streamflow measured at medium g gages (NWIS), on a log-scale. Red line denotes 1:1

Code

# dev.off()

2.10.2 Compare asynchronous gages

For Spread Creek and Shields River, compare data from co-located EcoDrought and NWIS gages, with NON-overlapping periods of record. Note that streamflow from NWIS gages is about ~1 order of magnitude greater than what is measured at EcoD gages.

--- title: "Collate Data" format: html editor: visual --- ```{r setup, include = FALSE} #knitr::opts_chunk$set(fig.retina = 2) ``` Purpose: Collate EcoDrought streamflow and temperature data, povided by EcoD PIs/partners and NWIS Notes: * need to figure out how to embed shiny/interactive plots. Tried 'shinylive' but this does not work, and appears to be a known issue (instability) with the latest Quarto release (v1.6.32) + here is the shinylive demo <https://github.com/coatless-quarto/r-shinylive-demo> + see <https://github.com/coatless-quarto/r-shinylive-demo/issues/14> which suggests downgrading * need to address missing data issue in the Flathead...interpolate at hourly timescale? ```{r, include = FALSE} library(tidyverse) library(mapview) library(sf) library(zoo) library(dataRetrieval) library(nhdplusTools) library(ggpubr) library(fasstr) library(dygraphs) library(ggforce) library(smwrBase) library(shiny) library(rmarkdown) library(shiny) library(htmlwidgets) ``` ## Gather site information ```{r} # West Brook siteinfo_wb <- read_csv("C:/Users/jbaldock/OneDrive - DOI/Documents/USGS/EcoDrought/EcoDrought Working/Data/Raw data/Mass/MA_site_info.csv") %>% mutate(Station_No = factor(Station_No), Site_Name = factor(Site_Name)) %>% rename(site_id = Site_ID, station_no = Station_No, site_name = Site_Name, lat = Latitude_dec_deg, long = Longitude_dec_deg, elev_ft = Elevation_ft, area_sqmi = Drainage_Area_sqmi) %>% mutate(designation = "little", basin = "West Brook", region = "Mass") #%>% select(-c(elev_ft, area_sqmi)) # Shenandoah siteinfo_shen <- read_csv("C:/Users/jbaldock/OneDrive - DOI/Documents/USGS/EcoDrought/EcoDrought Working/Data/Raw data/Virg/VA_site_info.csv") %>% mutate(Station_No = factor(Station_No), Site_Name = factor(Site_Name)) %>% rename(site_id = Site_ID, station_no = Station_No, site_name = Site_Name, lat = Latitude_dec_deg, long = Longitude_dec_deg, elev_ft = Elevation_ft, area_sqmi = Drainage_Area_sqmi) %>% mutate(designation = ifelse(str_detect(site_id, "10FL"), "big", "little"), basin = str_sub(site_name, 1, str_length(site_name)-3), region = "Shen") #%>% select(-c(elev_ft, area_sqmi)) # Flathead/Muhlfeld siteinfo_flat <- read_csv("C:/Users/jbaldock/OneDrive - DOI/Documents/USGS/EcoDrought/EcoDrought Working/Data/Raw data/Flathead/Flathead_SiteInfo_UpdateOct25.csv") %>% select(basin, site_name, site_id, region, designation, lat, long) %>% rename(subbasin = basin) %>% mutate(basin = "Flathead", region = "Flat") %>% select(site_id, site_name, lat, long, designation, basin, subbasin, region) %>% filter(designation == "little") # GYA/Al-Chokhachy siteinfo_gya <- read_csv("C:/Users/jbaldock/OneDrive - DOI/Documents/USGS/EcoDrought/EcoDrought Working/Data/Raw data/Al-Chokhachy/Al-Chokhachy_sites.csv") %>% mutate(region = ifelse(basin == "Snake River", "Snake", "Shields"), designation = "little") %>% select(site_id, site_name, latitude, longitude, designation, basin, region) %>% rename(lat = latitude, long = longitude) # NWIS Medium/Big/Super G sites <- c("01169900", # South River at Conway, Massachusetts "13011500", # Pacific Creek, Snake River, Wyoming "06195600", # Shields River at Livingston, Montana "12355500", # North Fork Flathead River, Montana "10396000", # Donner Blitzen River near Frenchglen, Oregon # Medium G "12355347", # Big Creek (Flathead) "12355342", # Hallowat Creek (Flathead) "06192980", # Shields Rivera above Smith Creek (GYA) "06192900", # Dugout Creek Mouth (GYA) "13012475", # South Fork Spread Creek (GYA) "13012465", # Leidy Creek, lower (GYA) "01171100", # West Brook (Mass) "01171000", # Avery Brook (Mass) "424551118503200", # Fish Creek at DB confluence (Oreg) "424547118503500", # DB above Fish Creek (Oreg) "424325118495900", # DB near Burnt Car Spring (Oreg) "424003118453700", # Little Blitzen River (Oreg) "423830118453200", # Indian Creek (Oreg) "423815118453900" # DB above Indian Creek (Oreg) ) siteinfo_nwis <- tibble(readNWISsite(sites)[,c(2:3,7,8,20,30)]) # get site info names(siteinfo_nwis) <- c("station_no", "site_name", "lat", "long", "elev_ft", "area_sqmi") # rename columns siteinfo_nwis <- siteinfo_nwis %>% mutate(site_name = c("South River Conway NWIS", "Avery Broook NWIS", "West Brook NWIS", "Dugout Creek NWIS", "Shields River ab Smith NWIS", "Shields River nr Livingston NWIS", "Donner Blitzen River nr Frenchglen NWIS", "Hallowat Creek NWIS", "Big Creek NWIS", "North Fork Flathead River NWIS", "Pacific Creek at Moran NWIS", "Leidy Creek Mouth NWIS", "SF Spread Creek Lower NWIS", "Donner Blitzen ab Indian NWIS", "Indian Creek NWIS", "Little Blizten River NWIS", "Donner Blitzen nr Burnt Car NWIS", "Donner Blitzen ab Fish NWIS", "Fish Creek"), site_id = c("SRC", "AVB", "WBR", "DUG", "SRS", "SRL", "DBF", "HAL", "BIG", "NFF", "PCM", "LEI", "SFS", "DBI", "IND", "LBL", "DBB", "DBA", "FSH"), designation = c("big", "medium", "medium", "medium", "medium", "big", "big", "medium", "medium", "big", "big", "medium", "medium", "medium", "medium", "medium", "medium", "medium", "medium"), basin = c("West Brook", "West Brook", "West Brook", "Shields River", "Shields River", "Shields River", "Donner Blitzen", "Flathead", "Flathead", "Flathead", "Snake River", "Snake River", "Snake River", "Donner Blitzen", "Donner Blitzen", "Donner Blitzen", "Donner Blitzen", "Donner Blitzen", "Donner Blitzen"), region = c("Mass", "Mass", "Mass", "Shields", "Shields", "Shields", "Oreg", "Flat", "Flat", "Flat", "Snake", "Snake", "Snake","Oreg", "Oreg", "Oreg", "Oreg", "Oreg", "Oreg")) %>% select(site_id, site_name, lat, long, station_no, designation, basin, region, elev_ft, area_sqmi) #mapview(st_as_sf(siteinfo_nwis, coords = c("long", "lat"), crs = 4326)) # bind together, fill in ragged subbasin siteinfo <- bind_rows(siteinfo_wb, siteinfo_shen, siteinfo_flat, siteinfo_gya, siteinfo_nwis) siteinfo$subbasin[siteinfo$site_name == "Hallowat Creek NWIS"] <- "Big Creek" siteinfo$subbasin[siteinfo$site_name == "Big Creek NWIS"] <- "Big Creek" siteinfo <- siteinfo %>% mutate(subbasin = ifelse(is.na(subbasin), basin, subbasin)) # fix Shields River Valley Ranch site locations siteinfo$lat[siteinfo$site_id == "SH07"] <- siteinfo$lat[siteinfo$site_id == "SRS"] siteinfo$long[siteinfo$site_id == "SH07"] <- siteinfo$long[siteinfo$site_id == "SRS"] # add elevation and area variables (from watershed delineation) areafiles <- list.files("C:/Users/jbaldock/OneDrive - DOI/Documents/USGS/EcoDrought/EcoDrought Working/EcoDrought-Analysis/Watershed Delineation/Area and Elevation") arealist <- list() for (i in 1:length(areafiles)) { arealist[[i]] <- read_csv(paste("C:/Users/jbaldock/OneDrive - DOI/Documents/USGS/EcoDrought/EcoDrought Working/EcoDrought-Analysis/Watershed Delineation/Area and Elevation/", areafiles[i], sep = ""))} areaelev <- do.call(rbind, arealist) # how well do provided and delineation area/elevation match? # siteinfo %>% left_join(areaelev, by = "site_id") %>% ggplot() + geom_point(aes(x = area_sqmi.x, y = area_sqmi.y)) + geom_abline(intercept = 0, slope = 1) + facet_wrap(~basin, scales = "free") # test %>% ggplot() + geom_point(aes(x = elev_ft.x, y = elev_ft.y)) + geom_abline(intercept = 0, slope = 1) + facet_wrap(~basin, scales = "free") # add delineated variables siteinfo <- siteinfo %>% select(-c(area_sqmi, elev_ft)) %>% left_join(areaelev) # fix NF Flathead (no dem from Canada) siteinfo$area_sqmi[siteinfo$site_id == "NFF"] <- 1556 ``` Write and re-load site information ```{r} write_csv(siteinfo, "C:/Users/jbaldock/OneDrive - DOI/Documents/USGS/EcoDrought/EcoDrought Working/Data/EcoDrought_SiteInformation.csv") siteinfo <- read_csv("C:/Users/jbaldock/OneDrive - DOI/Documents/USGS/EcoDrought/EcoDrought Working/Data/EcoDrought_SiteInformation.csv") ``` View unique basin names ```{r} unique(siteinfo$basin) ``` View unique site names ```{r} unique(siteinfo$site_name) ``` Map sites ```{r} # convert to spatial object and view on map #| fig-cap: "Map of EcoDrought project locations" siteinfo_sp <- st_as_sf(siteinfo, coords = c("long", "lat"), crs = 4326) mapview(siteinfo_sp, zcol = "designation") ``` ## Load EcoD data ```{r} # West Brook dat_wb <- read_csv("C:/Users/jbaldock/OneDrive - DOI/Documents/USGS/EcoDrought/EcoDrought Working/Data/Raw data/Mass/EcoDrought Continuous_MA.csv") %>% mutate(Station_No = factor(Station_No), Site_Name = factor(Site_Name)) %>% rename(station_no = Station_No, site_name = Site_Name, datetime = DateTime_EST, height = GageHeight_Hobo_ft, flow = Discharge_Hobo_cfs, tempf = WaterTemperature_HOBO_DegF) %>% mutate(tempc = (tempf - 32)*(5/9)) %>% select(station_no, site_name, datetime, height, flow, tempc) %>% left_join(siteinfo %>% select(-station_no)) # Shenandoah dat_shen <- read_csv("C:/Users/jbaldock/OneDrive - DOI/Documents/USGS/EcoDrought/EcoDrought Working/Data/Raw data/Virg/EcoDrought_Continuous_VA.csv") %>% mutate(Station_No = factor(Station_No), Site_ID = factor(Site_ID), Discharge_Hobo_cfs = as.numeric(Discharge_Hobo_cfs)) %>% rename(station_no = Station_No, site_id = Site_ID, datetime = DateTime_EST, height = GageHeight_Hobo_ft, flow = Discharge_Hobo_cfs, tempf = WaterTemperature_HOBO_DegF) %>% mutate(tempc = (tempf - 32)*(5/9)) %>% select(station_no, site_id, datetime, height, flow, tempc) %>% left_join(siteinfo) # pull in Big G data separately (UVA long-term gage sites) dat_shen_uva_q <- read_csv("C:/Users/jbaldock/OneDrive - DOI/Documents/USGS/EcoDrought/EcoDrought Working/Data/Raw data/Virg/Shen_BigG_Discharge_hourly_UVA.csv") %>% rename(flow = cfs) dat_shen_uva_t <- read_csv("C:/Users/jbaldock/OneDrive - DOI/Documents/USGS/EcoDrought/EcoDrought Working/Data/Raw data/Virg/Shen_BigG_TempEtc_hourly_UVA.csv") %>% select(site_id, datetime, tempc_mean) %>% rename(tempc = tempc_mean) dat_shen_uva <- dat_shen_uva_q %>% left_join(dat_shen_uva_t) %>% left_join(siteinfo) # bind usgs and uva data dat_shen <- bind_rows(dat_shen, dat_shen_uva) # Flathead/Muhlfeld flatfiles <- list.files("C:/Users/jbaldock/OneDrive - DOI/Documents/USGS/EcoDrought/EcoDrought Working/Data/Raw data/Flathead/Export2/EMA/Continuous Data") flatlist <- list() for (i in 1:length(flatfiles)) { #print(flatfiles[i]) flatlist[[i]] <- read_csv(paste("C:/Users/jbaldock/OneDrive - DOI/Documents/USGS/EcoDrought/EcoDrought Working/Data/Raw data/Flathead/Export2/EMA/Continuous Data/", flatfiles[i], sep = "")) %>% mutate(DateTime = mdy_hm(DateTime, tz = "MST"), site_name = gsub("EMA.csv", "", flatfiles[i])) } dat_flat <- bind_rows(flatlist) %>% select(DateTime, GageHeightFT, DischargeCFS, TempF, TempC, site_name, DischargeReliability, TempReliability) %>% rename(datetime = DateTime, height = GageHeightFT, flow = DischargeCFS, tempf = TempF, tempc = TempC) %>% mutate(DischargeReliability = as_factor(DischargeReliability), TempReliability = as_factor(TempReliability)) %>% mutate(tempf = ifelse(is.na(tempf), (tempc * (9/5)) + 32, tempf), tempc = ifelse(is.na(tempc), (tempf - 32) * (5/9), tempc)) %>% left_join(siteinfo) %>% select(-tempf) # Greater Yellowstone/Al-Chokhachy gyafiles <- list.files("C:/Users/jbaldock/OneDrive - DOI/Documents/USGS/EcoDrought/EcoDrought Working/Data/Raw data/Al-Chokhachy/Al-Chokhachy data files") gyalist <- list() for (i in 1:length(gyafiles)) { #print(gyafiles[i]) gyalist[[i]] <- read_csv(paste("C:/Users/jbaldock/OneDrive - DOI/Documents/USGS/EcoDrought/EcoDrought Working/Data/Raw data/Al-Chokhachy/Al-Chokhachy data files/", gyafiles[i], sep = "")) %>% mutate(date = mdy(date), DateTime = ymd_hms(paste(date, time, sep = " "), tz = "MST"), discharge = as.numeric(discharge)*35.314666212661) } dat_gya <- bind_rows(gyalist) %>% select(DateTime, depth, discharge, temperature, location) %>% rename(datetime = DateTime, height = depth, flow = discharge, tempc = temperature, site_name = location) %>% filter(site_name != "EF Henrys") %>% # drop weird duplicate site/year? mutate(site_name = dplyr::recode(site_name, "EF Above Confluence" = "EF Duck Creek ab HF", "EF Below Confluence" = "EF Duck Creek be HF", "NF Spread Creek" = "NF Spread Creek Lower", "Upper NF Spread Creek" = "NF Spread Creek Upper", "SF Spread Creek" = "SF Spread Creek Lower", "Upper SF Spread Creek" = "SF Spread Creek Upper", "Shields River above Dugout Creek" = "Shields River ab Dugout", "Upper Leidy Creek" = "Leidy Creek Upper", "Leidy Creek" = "Leidy Creek Mouth", "Spread Creek" = "Spread Creek Dam", "Shields River above Smith Creek" = "Shields River Valley Ranch")) %>% left_join(siteinfo) %>% filter(tempc <= 100) ``` Bind EcoD hourly flow/temp data with siteinfo and write to file ```{r} # bind together dat <- bind_rows(dat_wb, dat_shen, dat_flat, dat_gya) # unique(dat$site_name) write_csv(dat, "C:/Users/jbaldock/OneDrive - DOI/Documents/USGS/EcoDrought/EcoDrought Working/Data/EcoDrought_FlowTempData_Raw.csv") dat <- read_csv("C:/Users/jbaldock/OneDrive - DOI/Documents/USGS/EcoDrought/EcoDrought Working/Data/EcoDrought_FlowTempData_Raw.csv") ``` Check unique designations ```{r} unique(dat$designation) ``` Check unique regions ```{r} unique(dat$region) ``` Check unique basins ```{r} unique(dat$basin) ``` Check unique subbasins ```{r} unique(dat$subbasin) ``` ### Inspect hourly data ```{r} mysites <- unlist(unique(dat %>% filter(basin == "Flathead") %>% select(site_name))) myplots <- list() for (i in 1:length(mysites)) { test <- dat %>% filter(site_name == mysites[i]) %>% select(datetime, flow, DischargeReliability) %>% spread(key = DischargeReliability, value = flow) %>% arrange(datetime) %>% mutate(mindiff = as.numeric(lead(datetime) - (datetime))) na_index <- which(test$mindiff > 60) ds_blank <- tibble(index = na_index + 0.5, datetime = NA, "0" = NA, "1" = NA) myplots[[i]] <- test %>% select(-mindiff) %>% rowid_to_column("index") %>% union(ds_blank) %>% arrange(index) %>% select(-index) %>% mutate(datetime = if_else(is.na(datetime), lag(datetime) + hours(1), datetime)) %>% dygraph() %>% dyRangeSelector() %>% dyAxis("y", label = "flow (cfs)") %>% dyOptions(fillGraph = TRUE, drawGrid = FALSE, axisLineWidth = 1.5) } ``` ::: panel-tabset #### BigCreekLower ```{r, echo=FALSE} myplots[[1]] ``` #### BigCreekMiddle ```{r, echo=FALSE} myplots[[2]] ``` #### BigCreekUpper ```{r, echo=FALSE} myplots[[3]] ``` #### CoalCreekHeadwaters ```{r, echo=FALSE} myplots[[4]] ``` #### CoalCreekLower ```{r, echo=FALSE} myplots[[5]] ``` #### CoalCreekMiddle ```{r, echo=FALSE} myplots[[6]] ``` #### CoalCreekNorth ```{r, echo=FALSE} myplots[[7]] ``` #### CycloneCreekLower ```{r, echo=FALSE} myplots[[8]] ``` #### CycloneCreekMidddle ```{r, echo=FALSE} myplots[[9]] ``` #### CycloneCreekUpper ```{r, echo=FALSE} myplots[[10]] ``` #### HallowattCreekLower ```{r, echo=FALSE} myplots[[11]] ``` #### LangfordCreekLower ```{r, echo=FALSE} myplots[[12]] ``` #### LangfordCreekUpper ```{r, echo=FALSE} myplots[[13]] ``` #### McGeeCreekLower ```{r, echo=FALSE} myplots[[14]] ``` #### McGeeCreekTrib ```{r, echo=FALSE} myplots[[15]] ``` #### McGeeCreekUpper ```{r, echo=FALSE} myplots[[16]] ``` #### MeadowCreek ```{r, echo=FALSE} myplots[[17]] ``` #### NicolaCreek ```{r, echo=FALSE} myplots[[18]] ``` #### SkookoleelCreek ```{r, echo=FALSE} myplots[[19]] ``` #### WernerCreek ```{r, echo=FALSE} myplots[[20]] ``` #### WoundedBuckCreek ```{r, echo=FALSE} myplots[[21]] ``` ::: ```{r} dat %>% filter(site_name == "CoalCreekLower") %>% select(datetime, flow, DischargeReliability) %>% spread(key = DischargeReliability, value = flow) %>% dygraph() %>% dyRangeSelector() %>% dyAxis("y", label = "flow (cfs)") %>% dyOptions(fillGraph = TRUE, drawGrid = FALSE, axisLineWidth = 1.5) ``` ## Load NWIS data ```{r} # extract daily mean discharge and temp data from USGS NWIS website dat_superg_nwis <- tibble(readNWISdv(siteNumbers = sites, parameterCd = c("00010", "00060"), startDate = "1980-01-01", endDate = Sys.Date(), statCd = c("00003", "00001", "00002")))[,-c(1,5,7,9,11)] names(dat_superg_nwis) <- c("station_no", "date", "tempc_max", "tempc_min", "tempc_mean", "flow_mean") dat_superg_nwis <- dat_superg_nwis %>% left_join(siteinfo) # Manually grab Donner Blitzen above Indian Creek at the original timescale # daily mean flow dropped from above query, perhaps b/c in 2020 flow measurements were made every minute dbabind <- tibble(readNWISdata(sites = "423815118453900", service = "uv", startDate = "1980-01-01", endDate = Sys.Date()))[,c(2,3,6)] dbabind_daily <- dbabind %>% group_by(site_no, date(dateTime)) %>% summarize(nobs = n(), flow_mean = mean(X_00060_00000)) %>% filter(nobs %in% c(96, 1440)) %>% select(c(1,2,4)) names(dbabind_daily) <- c("station_no", "date", "flow_mean") dbabind_daily <- dat_superg_nwis %>% filter(station_no == "423815118453900") %>% select(-flow_mean) %>% left_join(dbabind_daily) # bind DBabInd to reset of data dat_superg_nwis <- rbind(dat_superg_nwis %>% filter(station_no != "423815118453900"), dbabind_daily) ``` ```{r, fig.height=8, fig.width=10} #| fig-cap: "(log) Stream flow (cfs) for Big G NWIS gages" dat_superg_nwis %>% filter(designation == "big") %>% ggplot() + geom_line(aes(x = date, y = log(flow_mean))) + facet_wrap(~site_name) ``` ```{r, fig.height=8, fig.width=10} #| fig-cap: "(log) Stream flow (cfs) for Medium G NWIS gages" dat_superg_nwis %>% filter(designation == "medium") %>% ggplot() + geom_line(aes(x = date, y = log(flow_mean))) + facet_wrap(~site_name) ``` ## Bind data and calc. daily means ```{r} # daily flow/temp summaries dat_daily <- dat %>% mutate(date = as_date(datetime)) %>% group_by(station_no, site_name, site_id, basin, subbasin, region, lat, long, elev_ft, area_sqmi, designation, date) %>% summarize(disch_reli = max(DischargeReliability), temp_reli = max(TempReliability), flow_mean = mean(flow), flow_min = min(flow), flow_max = max(flow), tempc_mean = mean(tempc), tempc_min = min(tempc), tempc_max = max(tempc)) %>% arrange(region, basin, site_name, date) %>% ungroup() # cbind EcoD and NWIS datasets dat_daily <- bind_rows(dat_daily %>% select(-flow_min, -flow_max, -tempc_min, -tempc_max), dat_superg_nwis %>% select(-tempc_max, -tempc_min), dat_superg_nwis %>% filter(site_id == "SRL") %>% mutate(subbasin = "Duck Creek") %>% select(-tempc_max, -tempc_min)) # add missing dates dat_daily <- fill_missing_dates(dat_daily, dates = date, groups = site_name) ``` View daily mean time series data, for example, site_name = CoalCreekLower. Notice the many shorts gaps in the daily data. ```{r eval = FALSE, echo = FALSE, include = FALSE} #| fig-cap: "Mean daily stream flow (cfs) by site" #| standalone: true library(shiny) library(dygraphs) library(tidyverse) ddd <- read_csv("C:/Users/jbaldock/OneDrive - DOI/Documents/USGS/EcoDrought/EcoDrought Working/Data/EcoDrought_FlowTempData_DailyWeekly.csv") ui <- fluidPage( selectInput(inputId = "mysite", label = "site_name", choices = unique(ddd$site_name)), dygraphOutput("plotDy") ) server <- function(input, output) { flow_data <- reactive({ ddd %>% filter(site_name == input$mysite) %>% select(date, flow_mean) }) output$plotDy <- renderDygraph({ dygraph(flow_data()) %>% dyRangeSelector() %>% dyAxis("y", label = "Mean daily flow (cfs)") %>% dyOptions(fillGraph = TRUE, drawGrid = FALSE, axisLineWidth = 1.5) }) } shinyApp(ui = ui, server = server) ``` ```{r} library(dygraphs) dat_daily %>% filter(site_name == "CoalCreekLower") %>% select(date, flow_mean) %>% dygraph() %>% dyRangeSelector() %>% dyAxis("y", label = "Mean daily flow (cfs)") %>% dyOptions(fillGraph = TRUE, drawGrid = FALSE, axisLineWidth = 1.5) ``` ## Interpolate missing data Small periods of missing data (\<24 hours) become a problem when aggregating at the daily and weekly time scales. \*Note that currently this discovers and fills missing data at the daily time scale, but should be changed to interpolate at the original timescale of the raw data (e.g., hourly). ```{r} # explore data gaps mysites <- unique(dat_daily$site_name) mynas <- list() for (i in 1:length(mysites)) { mydisch <- unlist(dat_daily$flow_mean[dat_daily$site_name == mysites[i]]) runsna <- rle(is.na(mydisch)) mynas[[i]] <- tibble(site_name = mysites[i], run = runsna$lengths[runsna$values == TRUE]) } mynas <- do.call(rbind, mynas) # mynas %>% filter(run < 100) %>% ggplot() + geom_histogram(aes(x = run)) + facet_wrap_paginate(~site_name, nrow = 4, ncol = 4, page = 1) # mynas %>% filter(run < 100) %>% ggplot() + geom_histogram(aes(x = run)) + facet_wrap_paginate(~site_name, nrow = 4, ncol = 4, page = 2) # mynas %>% filter(run < 100) %>% ggplot() + geom_histogram(aes(x = run)) + facet_wrap_paginate(~site_name, nrow = 4, ncol = 4, page = 3) # mynas %>% filter(run < 100) %>% ggplot() + geom_histogram(aes(x = run)) + facet_wrap_paginate(~site_name, nrow = 4, ncol = 4, page = 4) ``` Most gaps are relatively short ```{r} #| fig-cap: "Distributiion of lengths of missing data (days)" mynas %>% ggplot() + geom_histogram(aes(x = run)) + xlab("Days") + ylab("Frequency") ``` Zoomed in... ```{r} #| fig-cap: "Distributiion of lengths of missing data (days)" mynas %>% filter(run < 30) %>% ggplot() + geom_histogram(aes(x = run)) + xlab("Days") + ylab("Frequency") ``` Considering just the short gaps, which are likely a function of logger malfunction or Aquarius export issues, which sites are problematic? Answer: Flathead ```{r} #| fig-cap: "Frequency of short (<40 days) data gaps by site" mynas %>% filter(run < 30) %>% ggplot() + geom_bar(aes(site_name))+ theme(axis.text.x = element_text(angle = 90, vjust = 0.5, hjust=1)) ``` Fill short gaps (<=14 days...2x smoothing period) using time series interpolation ```{r} # fill short gaps (<=14 days...2x smoothing period) using time series interpolation datalist <- list() for (i in 1:length(mysites)) { datalist[[i]] <- dat_daily %>% filter(site_name == mysites[i]) %>% mutate(flow_mean_filled = fillMissing(flow_mean, max.fill = 14, span = 100)) } # bind and check 1:1 dat_daily_fill <- do.call(rbind, datalist) # dat_daily_fill %>% ggplot() + geom_point(aes(x = flow_mean, y = flow_mean_filled)) + facet_wrap(~site_name, scales = "free") ``` Explore interpolated/filled time series relative to original (daily) data Again, CoalCreekLower as an example. Would like to replace with shiny/interactive app ```{r eval = FALSE, echo = FALSE, include = FALSE} # explore individual sites #| fig-cap: "Mean daily observed and interpolated stream flow (cfs) by site" #| standalone: true library(shiny) library(dygraphs) library(tidyverse) ddd <- read_csv("C:/Users/jbaldock/OneDrive - DOI/Documents/USGS/EcoDrought/EcoDrought Working/Data/EcoDrought_FlowTempData_DailyWeekly.csv") ui <- fluidPage( selectInput(inputId = "mysite", label = "site_name", choices = unique(ddd$site_name)), dygraphOutput("plotDy") ) server <- function(input, output) { flow_data <- reactive({ ddd %>% filter(site_name == input$mysite) %>% select(date, flow_mean, flow_mean_filled) }) output$plotDy <- renderDygraph({ dygraph(flow_data()) %>% dyRangeSelector() %>% dySeries("flow_mean", strokeWidth = 5, color = "black") %>% dySeries("flow_mean_filled", strokeWidth = 1, color = "red") %>% dyAxis("y", label = "Mean daily flow (cfs)") %>% dyOptions(fillGraph = TRUE, drawGrid = FALSE, axisLineWidth = 1.5) }) } shinyApp(ui = ui, server = server) ``` ```{r} library(dygraphs) dat_daily_fill %>% filter(site_name == "CoalCreekLower") %>% select(date, flow_mean, flow_mean_filled) %>% dygraph() %>% dyRangeSelector() %>% dySeries("flow_mean", strokeWidth = 5, color = "black") %>% dySeries("flow_mean_filled", strokeWidth = 1, color = "red") %>% dyAxis("y", label = "Mean daily flow (cfs)") %>% dyOptions(fillGraph = TRUE, drawGrid = FALSE, axisLineWidth = 1.5) ``` ## Calculate yield ```{r} # convert cfs and basin area to metric dat_daily_fill <- dat_daily_fill %>% mutate(flow_mean_cms = flow_mean*0.02831683199881, flow_mean_filled_cms = flow_mean_filled*0.02831683199881, area_sqkm = area_sqmi*2.58999) # sites sites <- unique(dat_daily_fill$site_name) # site-specific basin area in square km basinarea <- dat_daily_fill %>% filter(!is.na(site_id)) %>% group_by(site_name) %>% summarize(area_sqkm = unique(area_sqkm)) # calculate yield yield_list <- list() for (i in 1:length(sites)) { d <- dat_daily_fill %>% filter(site_name == sites[i]) ba <- unlist(basinarea %>% filter(site_name == sites[i]) %>% select(area_sqkm)) yield_list[[i]] <- add_daily_yield(data = d, values = flow_mean_cms, basin_area = as.numeric(ba)) %>% left_join(add_daily_yield(data = d, values = flow_mean_filled_cms, basin_area = as.numeric(ba)) %>% rename(Yield_filled_mm = Yield_mm)) } dat_daily_fill_wyield <- do.call(rbind, yield_list) ``` ## Calculate 7-day means ```{r} dat_daily_fill_wyield <- dat_daily_fill_wyield %>% group_by(site_name) %>% mutate(flow_mean_7 = rollapply(flow_mean, FUN = mean, width = 7, align = "center", fill = NA), flow_mean_filled_7 = rollapply(flow_mean_filled, FUN = mean, width = 7, align = "center", fill = NA), tempc_mean_7 = rollapply(tempc_mean, FUN = mean, width = 7, align = "center", fill = NA), Yield_mm_7 = rollapply(Yield_mm, FUN = mean, width = 7, align = "center", fill = NA), Yield_filled_mm_7 = rollapply(Yield_filled_mm, FUN = mean, width = 7, align = "center", fill = NA)) %>% ungroup() %>% filter(!is.na(site_id)) # # view flow # dat_daily_fill_wyield %>% filter(site_name == "Avery Brook") %>% select(date, flow_mean_filled, flow_mean_filled_7) %>% dygraph() %>% dyRangeSelector() %>% # dySeries("flow_mean_filled", strokeWidth = 5, color = "black") %>% dySeries("flow_mean_filled_7", strokeWidth = 1, color = "red") # # # view yield # dat_daily_fill_wyield %>% filter(site_name == "Avery Brook") %>% select(date, Yield_filled_mm, Yield_filled_mm_7) %>% dygraph() %>% dyRangeSelector() %>% # dySeries("Yield_filled_mm", strokeWidth = 5, color = "black") %>% dySeries("Yield_filled_mm_7", strokeWidth = 1, color = "red") # unique(dat_daily_fill_wyield$basin) # unique(dat_daily_fill_wyield$subbasin) # unique(dat_daily_fill_wyield$region) # unique(dat_daily_fill_wyield$disch_reli) # unique(dat_daily_fill_wyield$temp_reli) ``` View little and medium g time series data (7-day mean yield), by sub-basin. Use the handles below the x-axis to adjust the time frame. ```{r} mysubbasins <- unique(dat_daily_fill_wyield$subbasin)[-c(4,7,13)] myplots <- list() for (i in 1:length(mysubbasins)) { myplots[[i]] <- dat_daily_fill_wyield %>% filter(designation %in% c("little", "medium"), subbasin == mysubbasins[i]) %>% mutate(logYield = log(Yield_filled_mm_7)) %>% select(date, site_name, logYield) %>% spread(key = site_name, value = logYield) %>% dygraph() %>% dyRangeSelector() %>% dyAxis("y", label = "ln(Yield, mm)") } ``` ::: panel-tabset #### Big Creek ```{r, echo=FALSE} myplots[[1]] ``` #### Coal Creek ```{r, echo=FALSE} myplots[[2]] ``` #### McGee Creek ```{r, echo=FALSE} myplots[[3]] ``` #### West Brook ```{r, echo=FALSE} myplots[[4]] ``` #### Paine Run ```{r, echo=FALSE} myplots[[5]] ``` #### Staunton River ```{r, echo=FALSE} myplots[[6]] ``` #### Duck Creek ```{r, echo=FALSE} myplots[[7]] ``` #### Shields River ```{r, echo=FALSE} myplots[[8]] ``` #### Snake River ```{r, echo=FALSE} myplots[[9]] ``` #### Donner Blitzen ```{r, echo=FALSE} myplots[[10]] ``` ::: ## Write out and read data ```{r} # write out write_csv(dat_daily_fill_wyield, "C:/Users/jbaldock/OneDrive - DOI/Documents/USGS/EcoDrought/EcoDrought Working/Data/EcoDrought_FlowTempData_DailyWeekly.csv") dat_daily <- read_csv("C:/Users/jbaldock/OneDrive - DOI/Documents/USGS/EcoDrought/EcoDrought Working/Data/EcoDrought_FlowTempData_DailyWeekly.csv") ``` ## View data availability ```{r} # summarize data availability dat_summ <- dat_daily %>% filter(site_id != "MRN") %>% group_by(basin, date, designation) %>% summarize(numflow = sum(!is.na(flow_mean)), numtemp = sum(!is.na(tempc_mean))) %>% gather(type, avail, numflow:numtemp) %>% mutate(type2 = as.factor(paste(designation, type, sep = "_"))) %>% mutate(type3 = as.numeric(type2), avail = na_if(avail, 0)) %>% ungroup() %>% filter(!is.na(avail)) # levels(dat_summ$type2) # unique(dat_summ$basin) ``` ```{r, fig.height = 6, fig.width = 12} # plot all years #| fig-cap: "Data availability by basin, all years" myplot <- dat_summ %>% ggplot(aes(x = date, y = type3)) + geom_errorbarh(aes(xmax = date, xmin = date, color = avail), size = 0.001) + scale_color_continuous(trans = "reverse", na.value = "grey60") + scale_y_discrete(limits = c("Big G flow", "Big G temp", "Medium g flow", "Medium g temp", "Little g flow", "Little g temp")) + labs(colour = "Number \nof sites") + ylab("") + xlab("Date") + facet_wrap(~ basin, nrow = 4, labeller = as_labeller(c(`West Brook` = "West Brook: G = South River Conway (NWIS)", `Staunton River` = "Staunton River: G = Staunton River 10 (UVA)", `Paine Run` = "Paine Run: G = Paine Run 10 (UVA)", `Piney River` = "Piney River: G = Piney River 10 (UVA)", `Snake River` = "Snake River: G = Pacific Creek Moran (NWIS)", `Shields River` = "Shields River: G = Shields River Livingston (NWIS)", `Flathead` = "NF Flathead River: G = NF Flathead (NWIS)", `Donner Blitzen` = "Donner Blitzen River: G = DB Frenchglen (NWIS)", `Duck Creek` = "Duck Creek: G = Shields River Livingston (NWIS)"))) myplot # jpeg("C:/Users/jbaldock/OneDrive - DOI/Documents/USGS/EcoDrought/EcoDrought Working/EcoDrought-Analysis/Data Availability/DataAvailability_byBasin.jpg", height = 6, width = 12, units = "in", res = 500) # myplot # dev.off() ``` ```{r, fig.height = 6, fig.width = 12} # plot recent years #| fig-cap: "Data availability by basin, recent years" # jpeg("./Data Availability/DataAvailability_byBasin_recent.jpg", height = 6, width = 12, units = "in", res = 500) myplot <- dat_summ %>% filter(date >= "2018-10-01") %>% ggplot(aes(x = date, y = type3)) + geom_errorbarh(aes(xmax = date, xmin = date, color = avail), size = 0.001) + scale_color_continuous(trans = "reverse", na.value = "grey60") + scale_y_discrete(limits = c("Big G flow", "Big G temp", "Medium g flow", "Medium g temp", "Little g flow", "Little g temp")) + labs(colour = "Number \nof sites") + ylab("") + xlab("Date") + facet_wrap(~ basin, nrow = 4, labeller = as_labeller(c(`West Brook` = "West Brook: G = South River Conway (NWIS)", `Staunton River` = "Staunton River: G = Staunton River 10 (UVA)", `Paine Run` = "Paine Run: G = Paine Run 10 (UVA)", `Piney River` = "Piney River: G = Piney River 10 (UVA)", `Snake River` = "Snake River: G = Pacific Creek Moran (NWIS)", `Shields River` = "Shields River: G = Shields River Livingston (NWIS)", `Flathead` = "NF Flathead River: G = NF Flathead (NWIS)", `Donner Blitzen` = "Donner Blitzen River: G = DB Frenchglen (NWIS)", `Duck Creek` = "Duck Creek: G = Shields River Livingston (NWIS)"))) myplot ``` ## Compare co-located gages ### Compare synchronous gages Compare streamflow data from co-located EcoDrought and NWIS gages with overlapping periods of record ```{r, fig.height = 6, fig.width = 6} #| fig-cap: "Streamflow measured at little g gages (EcoDrought) as a function of streamflow measured at medium g gages (NWIS), on a log-scale. Red line denotes 1:1" # WEST BROOK p1 <- dat_daily %>% filter(site_name == "West Brook 0") %>% select(date, flow_mean_7) %>% rename(little = flow_mean_7) %>% left_join(dat_daily %>% filter(site_name == "West Brook NWIS") %>% select(date, flow_mean_7) %>% rename(medium = flow_mean_7)) %>% ggplot() + geom_point(aes(x = log(medium), y = log(little))) + geom_abline(intercept = 0, slope = 1, color = "red") p2 <- dat_daily %>% filter(site_name == "Avery Brook") %>% select(date, flow_mean_7) %>% rename(little = flow_mean_7) %>% left_join(dat_daily %>% filter(site_name == "Avery Broook NWIS") %>% select(date, flow_mean_7) %>% rename(medium = flow_mean_7)) %>% ggplot() + geom_point(aes(x = log(medium), y = log(little))) + geom_abline(intercept = 0, slope = 1, color = "red") # FLATHEAD p3 <- dat_daily %>% filter(site_name == "BigCreekMiddle") %>% select(date, flow_mean_7) %>% rename(little = flow_mean_7) %>% left_join(dat_daily %>% filter(site_name == "Big Creek NWIS") %>% select(date, flow_mean_7) %>% rename(medium = flow_mean_7)) %>% ggplot() + geom_point(aes(x = log(medium), y = log(little))) + geom_abline(intercept = 0, slope = 1, color = "red") # jpeg("./Data Availability/LittleMedium_Co-Located_Gages.jpg", height = 6, width = 6, units = "in", res = 500) ggarrange(p1, p2, p3, ncol = 2, nrow = 2, labels = c("West Brook 0", "Avery Brook", "Big Creek (Flathead)")) # dev.off() ``` ### Compare asynchronous gages For Spread Creek and Shields River, compare data from co-located EcoDrought and NWIS gages, with NON-overlapping periods of record. Note that streamflow from NWIS gages is about ~1 order of magnitude greater than what is measured at EcoD gages. ::: panel-tabset #### Leidy Creek ```{r, echo=FALSE} dat_daily %>% filter(site_name %in% c("Leidy Creek Mouth", "Leidy Creek Mouth NWIS")) %>% mutate(logflow = log(flow_mean_filled_7)) %>% select(date, site_name, logflow) %>% spread(key = site_name, value = logflow) %>% dygraph() %>% dyRangeSelector() %>% dyAxis("y", label = "ln(7-day mean flow, cfs)") ``` #### SF Spread Creek Lower ```{r, echo=FALSE} dat_daily %>% filter(site_name %in% c("SF Spread Creek Lower", "SF Spread Creek Lower NWIS")) %>% mutate(logflow = log(flow_mean_filled_7)) %>% select(date, site_name, logflow) %>% spread(key = site_name, value = logflow) %>% dygraph() %>% dyRangeSelector() %>% dyAxis("y", label = "ln(7-day mean flow, cfs)") ``` #### Dugout Creek ```{r, echo=FALSE} dat_daily %>% filter(site_name %in% c("Dugout Creek", "Dugout Creek NWIS")) %>% mutate(logflow = log(flow_mean_filled_7)) %>% select(date, site_name, logflow) %>% spread(key = site_name, value = logflow) %>% dygraph() %>% dyRangeSelector() %>% dyAxis("y", label = "ln(7-day mean flow, cfs)") ``` #### Sheilds Valley Ranch ```{r, echo=FALSE} dat_daily %>% filter(site_name %in% c("Shields River Valley Ranch", "Shields River ab Smith NWIS")) %>% mutate(logflow = log(flow_mean_filled_7)) %>% select(date, site_name, logflow) %>% spread(key = site_name, value = logflow) %>% dygraph() %>% dyRangeSelector() %>% dyAxis("y", label = "ln(7-day mean flow, cfs)") ``` :::